- Email: [email protected]
Histological description of tooth formation in adult Eretmodus cf. cyanostictus (Teleostei, Cichlidae)
Archives of Oral Biology (2005) 50, 635—643
www.intl.elsevierhealth.com/journals/arob
Histological description of tooth formation in adult Eretmodus cf. cyanostictus (Teleostei, Cichlidae) E. Vandervennet a,*, A. Huysseune b,1 a
Royal Belgian Institute of Natural Sciences, Vetebrate Department, Vautierstraat 29, B-1000 Brussel, Belgium b Ghent University, Department of Biology, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium Accepted 5 December 2004
KEYWORDS Eretmodini; Cichlidae; Tooth development; Replacement tooth; Enameloid
Summary The Eretmodini, a tribe of closely related cichlids (Teleostei, Cichlidae) originating from Lake Tanganyika, possess oral tooth shapes ranging from conical (in Tanganicodus) over cylindrical (in Spathodus) to spatulate (in Eretmodus). Prior to a study aiming to understand how these distinctly different tooth shapes can be acquired in such closely related taxa, a detailed histological study was required of tooth formation in a representative of the eretmodines. Here, we present a histological description of replacement tooth development in Eretmodus cf. cyanostictus. Using light-microscopic observations on semithin as well as on ground sections, microradiographs and stereo-microscopic observations of both alizarine red S stained and unstained jaws we can conclude that tooth formation in adult E. cf. cyanostictus roughly corresponds with what is known for teleost tooth development in general. Remarkable features include the localization and shape of the epithelial downgrowth, the transient presence of a layer intermediate between inner dental epithelium (IDE) and outer dental epithelium (ODE), the asymmetric shape of the enamel organ, the fact that the pulp cavity recedes in front of the forming enameloid during enameloid formation, and finally, the pattern of matrix mineralisation and maturation, and the presence of pigment in the enameloid. The observation that the enamel organ in adult E. cf. cyanostictus develops asymmetrically is significant for understanding tooth shape variation in the Eretmodini. # 2005 Elsevier Ltd. All rights reserved.
Introduction * Corresponding author. Tel.: +32 2 627 44 28; fax: +32 2 627 41 41. E-mail addresses: [email protected] (E. Vandervennet), [email protected] (A. Huysseune). 1 Tel.: +32 9 264 52 29; fax: +32 9 264 53 44.
Teeth are common features of most classes of vertebrates. Like most other organs, teeth result from reciprocal epithelio-mesenchymal interactions. In all known species, the development of a tooth follows a general pathway in which several stages
0003–9969/$ — see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.archoralbio.2004.12.001
636
can be recognised (initiation, morphogenesis, cytodifferentiation, matrix deposition, mineralisation and maturation, attachment). As teeth are sometimes the only fossil remains of a species, dental features (shape, number and arrangement) are important characters used in systematics to unravel phylogenetic relationships between (groups of) vertebrates.1,2 The Eretmodini are a tribe of closely related cichlid fishes (Teleostei, Cichlidae) originating from Lake Tanganyika in East Africa.3 The four nominal species of this tribe, Eretmodus cyanostictus Boulenger, 1898, Tanganicodus irsacae Poll, 1950, Spathodus erythrodon Boulenger, 1900 and Spathodus marlieri Poll, 1950, resemble each other morphologically to a large extent. Eretmodines are originally divided into three genera on the basis of a distinctive oral tooth shape: spatulate with a crown and neck region in Eretmodus, cylindrical in Spathodus and conical in Tanganicodus.3—5 Based on mitochondrial DNA analyses, Ru ¨ber et al.6 report the existence of six different genetic lineages within this tribe. Within these lineages the three major types of tooth shape appear to have evolved several times, independently of one another. Individuals from a nominal species are distributed over two or more genetic lineages, implying that they probably belong to separate species. For instance, E. cyanostictus is distributed over lineage C (E. cyanostictus) and A (E. cf. cyanostictus). The former represents the originally described E. cyanostictus species, whereas the latter represents a new species. In order to understand, from a developmental perspective, how such dramatically different tooth shapes are acquired within closely related taxa, it is necessary to study the way in which development of an individual tooth is achieved. In particular we wish to know how the enamel organ folds during the process of morphogenesis of the tooth germ and how deposition of the matrix proceeds. Indeed, in teleosts, the epithelio-mesenchymal interface (i.e. the inner surface of the enamel organ) foreshadows the future tooth shape7, due to the specific way in which the hypermineralised cap (the enameloid) is deposited in teleosts. In the literature on teleost tooth development, few papers cover the entire development of a tooth germ (i.e., from initiation to attachment).8 Most studies deal only with specific aspects of tooth development such as enameloid deposition9—12 or mineralisation.13 Huysseune and Sire14 describe in detail the development of (oral) first-generation teeth in the cichlid Hemichromis bimaculatus. However, it is known that the development of replacement teeth in adult teleosts differs in a number of
E. Vandervennet, A. Huysseune aspects from first-generation teeth.14,15 Therefore, and given that a unique tooth pattern is present in the Eretmodini,16 it was necessary to gain a detailed knowledge of the development of replacement teeth in eretmodines as a basis to study the mechanism that generates different (oral) tooth shapes in closely related species as well as within the life cycle of one species. Here we describe the development of replacement teeth in adult Eretmodus cf. cyanostictus using light microscopical observations on semithin as well as on ground sections, microradiographs and stereo-microscopic observations of both alizarin red S stained and unstained jaws. In addition, we used atomic absorption spectroscopy (AAS) to analyse the pigment observed in erupted teeth of Eretmodus cf. cyanostictus, and in particular to measure their iron content. It has been reported that iron is the main contributor to this pigment.17,18
Material and methods Nine adult specimens of E. cf. cyanostictus (standard length (SL) ranging from 37.0 to 52.5 mm) were collected from Lake Tanganyika (1996). Immediately after capture in the field, the animals were fixed and preserved in 80% ethanol. The dissected oral jaws (mandibles and premaxillaries) of the specimens were refixed in a mixture of 1.5% paraformaldehyde and 1.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) to which 0.1 M EDTA was added for decalcification. The jaws were next postfixed in OsO4, dehydrated in a graded ethanol series, and embedded in epon. Serial cross sections of 2 mm thickness were prepared using glass knives, stained with toluidine blue and observed with a light microscope. To prepare ground sections, the jaws (mandibles and premaxillaries) of an adult E. cf. cyanostictus were dehydrated in a graded ethanol series and embedded in methyl metacrylate. Ground sections of 150 mm thickness were prepared using a Buehler Isomet low speed saw. Microradiographic images of the ground sections were made on a sensitive film by X-ray (20 kV, 8 mA) exposure. To study the jaws in toto, they were dissected out of the head. Some of these were cleared in 1—2% KOH, stained with 0.1% alizarin red S in 0.5% KOH and stored in glycerine, others were studied without clearing or staining. Several tooth germs were dissected out of the jaws and used for observation and photography. To determine the total iron content of the enameloid and dentine part of eretmodine teeth samples, we used atomic absorption spectroscopy. Prior to
Replacement tooth development in adult Eretmodus cf. cyanostictus
analysis, the samples were acidified with concentrated HNO3 to 1%. Analysis was carried out on a Varian 800 Atomic Absorption Spectrophotometer with Zeeman background correction. Because of the technical difficulties encountered, we could analyse only five teeth (one whole functional tooth, the enameloid part of one functional tooth, the dentine part of one functional tooth, one unpigmented tooth germ and one orange-coloured membrane covering the enameloid of an unpigmented tooth germ).
Results Since, in fish, tooth replacement takes continuously place during lifetime, jaws of adult E. cf. cyanostictus showed a considerable number of tooth germs, in all stages of development, from initiation stage, up to attached, fully functional teeth. The distribution and localization of the tooth germs in the jaw was in accordance to the pattern described by Huysseune et al.16 Below, a description will be given of the development of replacement teeth in adult E. cf. cyanostictus.
Stage of initiation The first histomorphological indication of tooth formation in E. cf. cyanostictus that we were able to identify was a local downgrowth of the epithelium (Fig. 1). This downgrowth always budded off from the epithelium at the bottom of the mucosal crypt that surrounds the erupted part of a functional tooth. The epithelial strand in E. cf. cyanostictus did not develop straight downwards. Shortly behind the point of budding the strand bent so that it ran parallel to the oral epithelium for some distance. Then the downgrowth penetrated the jawbone and entered the mesenchyme of the medullary cavity to reach the area where the future tooth germ was about to form. This epithelial strand persisted throughout the development of the tooth germ and connected the developing tooth germ to the mucosal surface.
637
(Fig. 2). At this stage, the ODE was composed of cuboidal to slightly cylindrical cells, whereas the cells of the IDE (the future ameloblasts) were already cylindrical. In several tooth germs, we observed several layers of elongated, spindleshaped cells squeezed between IDE and ODE along the prospective convex side of the enamel organ (Fig. 2). At the time the enamel organ formed, the mesenchyme about to become the dental papilla was already distinguishable from the mesenchyme that surrounded the tooth germ. Whereas the latter consisted of elongated spindle-shaped fibroblastlike cells, the mesenchymal cells of the prospective dental papilla were already clearly differentiated: the cells were more rounded, with a larger nucleus, and more cytoplasm. The cells facing the basal lamina (the future odontoblasts) had already taken a slightly polarized appearance (Fig. 2).
Stage of enameloid matrix formation The first matrix to be deposited was the unmineralised precursor of the enameloid. This matrix was a homogeneous, chromophilic substance, in which fibres or fibre bundles were hardly detectable, and which lacked any cellular prolongations emerging from the odontoblasts (Fig. 3). Deposition of enameloid matrix proceeded in an asymmetrical way. The distal side had a delay in comparison with the mesial side. At this stage, the enamel organ was highly differentiated. The ameloblasts facing this unmineralised matrix were extremely tall cells, of slightly varying height, thus resulting in a wavy appearance of the outline of the ameloblast layer. The nuclei of the ameloblasts were situated in the proximal end of the cell (i.e., turned away from the tooth surface). In contrast, the ODE cells were cuboidal and formed a thin layer that followed the wavy outline of the IDE. Many capillaries adjoined the ODE. At this stage, they started to become entrapped within the folds of the enamel organ. The spindle-shaped cells between IDE and ODE were no longer observed.
Stage of morphogenesis
Stage of enameloid matrix mineralisation and maturation
Where the tooth germ was about to develop, the deep end of the epithelial strand next folded into what is called the enamel organ. From the earliest stages onwards, this enamel organ consisted of two layers; the outer dental epithelium (ODE) and the inner dental epithelium (IDE). Initially, the enamel organ extended deeper into the medullary cavity on the future convex and mesial side of the tooth
The enameloid matrix was deposited centripetally, i.e., from the basal lamina inwards. It formed a blade with a convex and a concave surface. The mineralisation of the enameloid matrix started before its deposition was completed. Fully mineralised enameloid hardly showed any evidence of remains of the organic matrix and did not take up any stain.
638
We observed several stages in the development of the enameloid matrix, possibly reflecting the degree of mineralisation. In the youngest tooth germs with ongoing maturation, the fully mineralised matrix was present at the convex (labial) side only (Fig. 4). Maturation then progressed on both sides of the enameloid blade, but left a thin zone of less mineralised matrix in the centre (Fig. 5). The enamel organ was extensively folded during the mineralisation and maturation stages. Whereas
E. Vandervennet, A. Huysseune
the IDE remained in contact with the tooth surface throughout all stages of odontogenesis, folds of the ODE surrounded blood capillaries. The folding of the enamel organ did not occur to the same extent along both sides of the tooth. Indeed, folding was always more pronounced on the side with more advanced matrix maturation, whereas the side facing the less mineralised matrix showed less pronounced folding. When the enamel organ first became moderately folded, dense chromophilic deposits appeared in the
Figures 1—8 E. cf. cyanostictus: toluidine blue stained semithin sections of tooth germs in different stages of development. (Fig. 1) Initiation stage. Note the epithelial downgrowth (black arrowhead) originating from the reduced enamel organ (black asterisk) of a functional tooth (white asterisk). (Fig. 2) Morphogenesis stage. The enamel organ is composed of an inner dental epithelium (white arrowhead) and an outer dental epithelium (black arrowhead). The enamel organ penetrates deeper into the mesenchyme on the future convex side of the tooth. Along this side the two layers seem to be separated by an epithelial layer (black asterisk). The mesenchymal cells of the prospective dental papilla are indicated by a white asterisk. (Fig. 3) Tooth germ in stage of enameloid matrix deposition (white asterisk); the ameloblasts (black asterisk) are extremely tall polarised cells. Note the presence of capillaries along the outer boundary of the enamel organ (arrows). (Fig. 4) Stage of ongoing enameloid mineralisation. The fully mineralised enameloid (black arrowhead) hardly takes up any stain; only traces of organic matrix are left. The less mineralised enameloid is indicated by the white asterisk. (Fig. 5) Tooth germ in a further stage of enameloid mineralisation. Fully mineralised enameloid (black arrowheads) surrounds a central zone of less mineralised enameloid (black asterisk). A white asterisk indicates the dentine matrix. (Fig. 6) Detail of ameloblasts, showing dense chromophilic deposits (black arrowhead) accumulating against the immature mineralised enameloid matrix (black asterisk). (Fig. 7) Cross-sectioned tooth germ showing details of the dentine matrix. Note a pseudo-epithelial layer of odontoblasts (black arrowhead) lining the pulp cavity (black asterisk). Fibre bundles in the dentine are indicated by white arrowhead. (Fig. 8) Longitudinal section through a newly attached tooth. Black arrowheads indicate the unmineralised borderline between dentine base and attachment bone. Small black arrowheads point to osteoblasts lining the surface of the attachment bone. The tip of the cervical loop is highlighted by black asterisks.
Replacement tooth development in adult Eretmodus cf. cyanostictus
639
Fig. 1. (Continued).
ameloblasts opposite the slightly mineralised matrix. These deposits accumulated in the distal part of the ameloblasts, facing the enameloid (Fig. 6). Next, when the enameloid matrix matured, and the enamel organ was elaborately folded, such deposits were found more dispersed in the ameloblasts (Fig. 5); they continued to be present during the whole maturation stage of the enameloid. In some tooth germs, such deposits were also found either at the boundary line between fully mineralised and less mineralised enameloid matrix, or arranged as two lines within the zone of mineralised enameloid (Fig. 5).
Stage of dentine deposition and mineralisation Once deposition of the unmineralised enameloid matrix was finished, the deposition of the dentine matrix started. It is noteworthy that, as long as enameloid was being deposited, the dental papilla steadily receded in front of the forming enameloid, so that the enameloid cap was deposited between two opposite layers of IDE. In contrast, dentine was deposited at the contact area between the dental papilla and the enameloid. Contrary to the enameloid matrix, to which it superficially resembled, the dentine was characterized by the presence of dis-
tinct fibre bundles (Fig. 7). Although cellular prolongations emanating from the odontoblasts were not obvious in the early phase of dentine deposition, a thicker layer of dentine often showed very distinct and long odontoblastic processes. During dentine deposition, the odontoblasts formed a palisade of high columnar cells arranged in a pseudo-epithelial manner against the matrix (Fig. 7). The rest of the dental papilla consisted of a loose network of mesenchymal cells. Around the forming dentine shaft the enamel organ was no longer folded, capillaries were no longer closely associated with the ODE and the cells of the IDE were again low columnar cells. Deposition of the dentine shaft had usually progressed further along one side of the tooth, indicating that its deposition, like the enameloid deposition, proceeded in an asymmetrical way. In contrast to the enameloid, mineralised dentine retained its ability to stain.
Stage of attachment Beneath the base of the dentine shaft, which ended at the level of the cervical loop tip, an annular bone of attachment was deposited (Fig. 8). The tooth base and the attachment bone were ligamentously connected but did not fuse. The inner surface of the
640
E. Vandervennet, A. Huysseune
attachment bone was lined by osteoblasts, which lied in the prolongation of the odontoblasts of the dentine base. Numerous capillaries invaded the pulp cavity.
Table 1 Iron content expressed as absolute values for a complete functional tooth, its constituting parts (enameloid and dentine, respectively), an unpigmented tooth germ and an isolated membrane (enamel organ?), for E. cf. cyanostictus.
Ground sections
Sample
Fe (mg/tooth)
Functional tooth Enameloid Dentine Non-pigmented tooth Pigmented membrane
0.927 0.597 0.094 0.030 0.087
In ground sections observed in transmitted light, all tooth germs appeared to be fully mineralised, with no evidence for a distinction between mineralised and unmineralised enameloid matrix. This was confirmed by comparing the ground sections with their respective microradiographs. Parts of the dentine shaft near the tooth base remained unmineralised.
Iron deposition in enameloid of E. cf. cyanostictus In live and preserved specimens of E. cf. cyanostictus, the enameloid of the functional teeth had a deep orange colour. On semithin sections, an orange-coloured pigment could be observed in the enameloid of (erupted) functional teeth, but never in the tooth germs. In dissected tooth germs of alizarin red S-stained jaws, the enameloid cap showed a colour gradient which corresponded to the developmental age of the tooth germ: from dark pink (intense alizarin red S uptake) in tooth germs with incipient dentine formation, to dark orange (no alizarin red S uptake) in fully developed teeth (Fig. 9). Remarkably, in tooth germs with starting dentine formation, a thin membrane could be peeled off from the enameloid cap of the tooth germ, which appeared to be loaded with alizarin red S positive granules (Fig. 10). In unstained jaws this membrane was loaded with an orange pigment. Such a pigmented membrane was only observed around tooth germs with an unpigmented (white) or slightly pigmented (light orange) enameloid cap,
but not around tooth germs that were more heavily pigmented (dark orange). The results of the iron content analysis of different (parts of) teeth are shown in Table 1. The enameloid contained more iron than the dentine. A functional (pigmented) tooth had a higher iron content than a (almost) fully developed tooth lacking this orange pigment. The iron content of the pigmented membrane, which could be peeled off the enameloid cap of an E. cf. cyanostictus tooth germ, seemed to be low. It should be noted however, that the membrane for which the iron content was measured, only represents a fraction of the weight of the (parts of the) tooth used.
Discussion The way in which the histological development of an E. cf. cyanostictus tooth proceeds complies roughly with what is known for tooth development in other cichlids, or teleost tooth development in general.19 Some remarkable features observed in eretmodine tooth development nevertheless deserve particular attention and will be discussed below. They include the localization and shape of the epithelial down-
Figures 9 and 10 (Fig. 9) Series of dissected tooth germs, prepared from alizarin red S stained jaws of E. cf. cyanostictus, arranged according to an increasing degree of development and mineralisation, from incipient dentin formation (left) to fully formed tooth (right). (Fig. 10) Alizarin red S stained membrane peeled off the enameloid cap of a tooth crown in adult E. cf. cyanostictus similar in appearance to the left tooth germ shown in Fig. 9.
Replacement tooth development in adult Eretmodus cf. cyanostictus
growth, the transient presence of a layer intermediate between IDE and ODE, the close association of capillaries with the enamel organ, the presence of pigment, the asymmetric shape of the enamel organ, the pattern of mineralisation, and finally, the steady withdrawal of the pulp cavity. The tooth germs always budded off from the epithelium at the bottom of the mucosal crypt surrounding the erupted predecessor. We never observed an epithelial strand deriving directly from the oral epithelium. It was recently proposed that this localization, which is also observed in other teleosts, might be related to the presence of a stem cell niche possibly required for the formation of replacement teeth.20 Remarkably, the epithelial strand in E. cf. cyanostictus followed a wavy course prior to penetrating the medullary cavity of the dentary bone. This may be related to the arrangement of the teeth that results in the particular replacement pattern that characterizes eretmodines. Indeed, in eretmodines, the dentition is composed of subsequent oblique rows of teeth, called tooth groups, and is characterized by betweengroup replacement.16 This pattern implies the simultaneous functionality of different teeth belonging to the same tooth family. Hence, predecessor and replacement tooth overlap in their time window of functionality, and therefore the replacement tooth cannot attach at the same position as its predecessor. The replacement tooth needs to develop at some distance labial from its predecessor, a situation possibly enabled by the displacement of the downgrowth of the epithelial strand connecting predecessor and replacement tooth. In most Actinopterygii studied, the enamel organ consists of two layers, the IDE and ODE, which are immediately superimposed onto each other.19 There are nevertheless reports of intermediate cell layers between the IDE and the ODE.10,21—23 The composition of the intermediate layers shows variations among different groups of fish. Sasagawa10 reports that a number of epithelial cells, resembling those of the stellate reticulum in higher vertebrates, are temporarily present between IDE and ODE cells in the cichlid Tilapia nilotica (Oreochromis niloticus) at the middle stage of formation of enameloid matrix. We show here that in E. cf. cyanostictus too, such group of cells was temporarily present during enameloid matrix formation. It seems likely that these cells are involved in the mineralisation process. In E. cf. cyanostictus, as described for other Actinopterygii,13,24,25 capillaries became entrapped in the folds of the enamel organ during the mineralisation and maturation stages, thereby creating a highly folded outer surface of the enamel organ.
641
During these stages, the dental epithelial cells (both IDE and ODE) are involved in the removal of the organic matrix from the enameloid and in the process of mineralisation.26 The ODE cells themselves are involved in the active transport of inorganic ions (calcium, fluoride, iron, . . .) from the capillaries to the IDE.26 The iron is concentrated by the IDE cells as ferritin. This ferritin is digested in ferritin-containing vesicles to release the iron pigment to the enamel.17 In Tilapia the concentration of iron in the IDE cells starts to increase with the appearance of pigment in the same epithelial cells.17 In dissected but unstained jaws of adult E. cf. cyanostictus, we observed a membrane loaded with an orange pigment covering the enameloid cap of tooth germs in the phase of dentine matrix formation (thus, during enameloid mineralisation and maturation stages). The pigmented membrane was only present on teeth that still had an unpigmented or slightly pigmented cap. Later, the enameloid cap became deep orange and a pigmented membrane was no longer detectable. We suspect that this membrane represents the (inner) dental epithelium and that the pigment is a precursor of the iron-containing pigment to be deposited in the enameloid. Iron analysis of eretmodine teeth (Table 1) supports this view. Depending on when the iron deposition into the developing enameloid starts, fish teeth either have iron distributed throughout the entire enameloid layer (in the case of an early start), or they possess an iron-rich surface layer (in the case of a late start).27 In Cichlidae, iron is deposited throughout the entire enameloid layer with the highest concentration at the surface or middle layers.27,28 In functional teeth of adult E. cf. cyanostictus, an orange pigment was dispersed throughout the entire enameloid layer. In light microscopic sections, dense chromophilic deposits were located first in the distal part of the ameloblasts facing slightly mineralised enameloid. In later stages these deposits were dispersed in the ameloblasts (Figs. 5 and 6). These deposits were present from the beginning of the mineralisation stage. Later they also appeared in the enameloid matrix. The order of appearance and localisation of these chromophilic deposits fits our observations for the orange iron pigment. Therefore we suggest that the chromophilic depositions, visualized with toluidine blue dye, correspond to the pigment seen around and in the teeth of unstained jaws, as described above. By the time the enameloid matrix deposition had started, the enamel organ extended deeper into the jaw medullary cavity on the prospective convex surface and on the prospective mesial side of the enameloid cap. Later, the invasion of capillaries, and thus
642
the formation of folds within the enamel organ, started on the convex surface of the tooth as well. Also a delay in matrix deposition on one side of the tooth was observed. Together, these observations suggest that tooth development proceeds asymmetrically. These findings are highly relevant for our interpretation of the acquisition of tooth shape differences between species of eretmodines. Microradiographs of ground sections did not reveal differences in the degree of mineralisation of the enameloid cap. The staining difference observed in the enameloid matrix in toluidine blue stained sections is therefore likely not due to a difference between unmineralised and mineralised matrix, but probably reflects differences in the maturation of the enameloid. Mineralisation of the enameloid starts before its matrix deposition is completed and proceeds in a way which contrasts with other species for which descriptions are available (Cichlasoma,13 Tilapia,29 Hoplognathus30,31,32). In these species, enameloid mineralisation starts when the future outline of the enameloid matrix is completed. In the species listed above, mineralisation always starts at the enameloid-dentine boundary. In Hoplognathus fasciatus, further completion of mineralisation takes place from the surface of enameloid inwards, while in Tilapia nilotica and in Cichlasoma cyanoguttatum, both cichlids, the mineralisation spreads out centrifugally in the cap enameloid. The remarkably different pattern of mineralisation in E. cf. cyanostictus is possibly related to the steady withdrawal of the dental papilla, resulting in the deposition of the enameloid matrix between the two opposing IDE layers, and leaving no pulp cavity in the fully formed enameloid cap. In conclusion, our observations have revealed that the development of an adult E. cf. cyanostictus tooth corresponds to a large extent to what is known for teleost oral tooth development in general (place of initiation of replacement tooth, downgrowth of an epithelial strand, development of the enamel organ within the medullary cavity of the bone). Some features appeared to be related particularly to the eretmodine pattern of the dentition (direction of downgrowth of the epithelial strand), and still others to the particular shape of the teeth (features of the enamel organ, pattern of enameloid mineralisation and maturation and withdrawal of the dental papilla). This description has laid the basis for a detailed study of the enamel organs in the closely related species of eretmodines, aiming to unravel how, from a developmental viewpoint, distinctly different tooth shapes can be generated in such closely related species. The observation of an asymmetrical enamel organ is significant in this respect. Compar-
E. Vandervennet, A. Huysseune
ison with the conical shape of adult T. cf. irsacae and juvenile E. cf. cyanostictus, should allow us to present a model that shows how the folding of the enamel organ changes between conical and spatulate teeth.
Acknowledgments The authors thank Dr. J.Y. Sire and Mrs. M.H. Deschamp (CNRS, Universite ´ Paris 6) for the microradiographs and Prof. Dr. R. Blust and Drs. I. Komjarova (Laboratory for Ecophysiology, Biochemistry and Toxicology, University of Antwerp) for the iron analysis. The research performed by Els Vandervennet is supported by a grant from the Belgian Federal Science Policy Office. Our research further benefited by a grant from the ‘Fonds voor wetenschappelijk onderzoek — Vlaanderen’ (3G010999) and the ‘Bijzonder onderzoeksfonds’ from Ghent University (011B2396).
References 1. Greenwood PH. The haplochromine fishes of the east african lakes. Mu ¨nchen: Kraus International Publications; 1981 (p. 1—839). 2. Thenius E. Za ¨hne und Gebiss der Sa ¨ugetiere. In: Niethammer J, Schliemann H, Starck D, editors. Handbuch der zoologie. Berlin: Walter de Gruyter; 1989. 3. Poll M. Classification des cichlidae du Lac Tanganika: Tribus, ´m Sci 1986;45:1—163. genres et espe `ces. Acad Roy Belg-Me 4. Liem KF. Modulatory multiplicity in the feeding mechanism in cichlid fishes, as exemplified by the invertebrate pickers of Lake Tanganyika. J Zool Lond 1979;189:93—125. 5. Wautier K, Huysseune A, Verheyen E. Tooth shape differences analyzed by biometric and morphometric approaches; a case study on two morphologically very similar lacustrine cichlid species. Connect Tissue Res 2002;43:103—8. 6. Ru ¨ber L, Verheyen E, Meyer A. Replicated evolution of trophic specialisations in an endemic cichlid fish lineage from Lake Tanganyika. Proc Natl Acad Sci USA 1999;96:10230—5. 7. Shellis RP. The role of the IDE in the formation of the teeth in fish. In: Butler PM, Joysey KA, editors. Development, function and evolution of teeth. London: London Academic Press; p. 31—42. 8. Huysseune A, Sire JY. Structure and development of the teeth in three armoured catfish, Corydoras aeneus, C. arcuatus and Hoplosternum littorale (Siluriformes, Callichthyidae). Acta Zool 1997;78:69—84. 9. Shellis RP, Miles AE... Observations with the electron microscope on enameloid formation in the common eel (Anguilla anguilla; Teleostei). Proc R Soc Lond (Biol) 1976;194:253— 69. 10. Sasagawa I. Fine structure of tooth germs during the formation of enameloid matrix in Tilapia nilotica, a teleost fish. Archs Oral Biol 1995;40:801—14. 11. Prostak K, Skobe Z. Ultrastructure of the dental epithelium and odontoblasts during enameloid matrix deposition in cichlid teeth. J Morphol 1986;187:159—72.
Replacement tooth development in adult Eretmodus cf. cyanostictus
12. Prostak K, Seifert P, Skobe Z. Enameloid formation in two tetraodontiform fish species with high and low fluoride contents in enameloid. Arch Oral Biol 1993;38:1031—44. 13. Prostak K, Skobe Z. Ultrastructure of the dental epithelium during enameloid mineralisation in a teleost fish. Cichlasoma cyanoguttatum. Arch Oral Biol 1986;31:73—85. 14. Huysseune A, Sire JY. Structure and development of firstgeneration teeth in the cichlid Hemichromis bimaculatus (Teleostei, Cichlidae). Tissue and Cell 1997;29:679—97. 15. Sire JY, Davit-Beal T, Delgado S, Van Der Heyden C, Huysseune A. First-generation teeth in nonmammalian lineages: evidence for a conserved ancestral character. Microsc Res Tech 2002;59:408—34. 16. Huysseune A, Ru ¨ber L, Verheyen E. A single tooth replacement pattern generates diversity in the dentition in cichlids of the tribe Eretmodini, endemic to Lake Tanganyika (Teleostei: Cichlidae). Belg J Zool 1999;129:157—74. 17. Sasagawa I. Iron accumulation in the dental epithelial cells of Tilapia nilotica, a teleost. ACBTE 1993;3:19—28. 18. Garant P. Observations on the ultrastructure of the ectodermal component during odontogenesis in Helostoma temmincki. Anat Rec 1970;166:167—88. 19. Huysseune A, Sire JY. Evolution of patterns and processes in teeth and tooth-related tissues in nonmammalian vertebrates. Eur J Oral Sci 1998;106:437—781. 20. Huysseune A, Thesleff I. Continuous tooth replacement: the possible involvement of epithelial stem cells. BioEssays 2004;26:665—71. 21. Kerr T. Development and structure of some actinopterygian and urodele teeth. Proc R Soc Lond B 1960;194:253— 69. 22. Peyer B. Comparative odontology. Chicago: University of Chicago Press; 1968 p. 1—347. 23. Ishiyama M, Inage T, Simokawa H. A immunocytochemical study of amelogenin proteins in the developing tooth enamel of the garpike, Lepisosteus oculatus (Holostei, Actinopterygii). Arch Histol Cytol 1999;62:191—7.
643
24. Sasagawa I, Ishiyama M. Fine structure and Ca-ATPase activity of the stratum intermedium cells during ondontogenesis in gars, Lepisosteus Actinopterygii. Connect Tissue Res 2002; 43:505—8. 25. Sasagawa I, Ishiyama M. The features of enameloid formation during odontogenesis in teleosts. In: Mayhall JT, Heikkinen T, editors. Dental Morphology 1998 Proceedings of the 11th international symposium on dental morphology Oulu University Press; 1999. p. 285—92. 26. Sasagawa I. Fine structure of cap enameloid and of the dental epithelial cells during enameloid mineralisation and early maturation stages in the tilapia, a teleost. J Anat 1997;190: 589—600. 27. Suga S, Taki Y. Fluoride and iron concentrations in the teeth of primitive bony fishes (Osteichthyes). In: Suga S, Watabe N, editors. Hard tissue mineralization and demineralization. Tokyo, Hong Kong: Springer Verlag; 1992. p. 17—32. 28. Suga S, Taki Y, Wada K, Ogawa M. Evolution of fluoride and iron concentrations in the enameloid of fish teeth. In: Suga S, Nakahara H, editors. Mechanisms and phylogeny of mineralization in biological systems. Tokyo: Springer Verlag; 1991. p. 439—46. 29. Sasagawa I. Features of the cap enameloid in the pharyngeal teeth of tilapia, a teleost. ACBTE 1995;4:7—14. 30. Isokawa S, Tsubouchi M, Aoki K, Imai M, Kawai A, Tsuchida S. Studies on the developing enameloid of a fish (Hoplognathus fasciatus) I. Mineralization pattern of enameloid matrix. J Nihon Univ Sch Dent 1970;12:43—9. 31. Isokawa S, Yoshida M, Ikeda K, Yamaguchi K, Chikazawa N, Fujioka S. Tooth enameloid formation of a fish (Hoplognathus fasciatus) chronological relationship of matrix formation and mineralization of the enameloid and dentin. J Nihon Univ Sch Dent 1972;14:1—11. 32. Inoue Y, Uchino S, Asano T, Hasegawa S, Fujioka Y. The mineralisation pattern of developing fish tooth enameloid using X-ray historadiography and radioactive tracer (45Ca). J Nihon Univ Sch Dent 1972;15:52—7.
www.intl.elsevierhealth.com/journals/arob
Histological description of tooth formation in adult Eretmodus cf. cyanostictus (Teleostei, Cichlidae) E. Vandervennet a,*, A. Huysseune b,1 a
Royal Belgian Institute of Natural Sciences, Vetebrate Department, Vautierstraat 29, B-1000 Brussel, Belgium b Ghent University, Department of Biology, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium Accepted 5 December 2004
KEYWORDS Eretmodini; Cichlidae; Tooth development; Replacement tooth; Enameloid
Summary The Eretmodini, a tribe of closely related cichlids (Teleostei, Cichlidae) originating from Lake Tanganyika, possess oral tooth shapes ranging from conical (in Tanganicodus) over cylindrical (in Spathodus) to spatulate (in Eretmodus). Prior to a study aiming to understand how these distinctly different tooth shapes can be acquired in such closely related taxa, a detailed histological study was required of tooth formation in a representative of the eretmodines. Here, we present a histological description of replacement tooth development in Eretmodus cf. cyanostictus. Using light-microscopic observations on semithin as well as on ground sections, microradiographs and stereo-microscopic observations of both alizarine red S stained and unstained jaws we can conclude that tooth formation in adult E. cf. cyanostictus roughly corresponds with what is known for teleost tooth development in general. Remarkable features include the localization and shape of the epithelial downgrowth, the transient presence of a layer intermediate between inner dental epithelium (IDE) and outer dental epithelium (ODE), the asymmetric shape of the enamel organ, the fact that the pulp cavity recedes in front of the forming enameloid during enameloid formation, and finally, the pattern of matrix mineralisation and maturation, and the presence of pigment in the enameloid. The observation that the enamel organ in adult E. cf. cyanostictus develops asymmetrically is significant for understanding tooth shape variation in the Eretmodini. # 2005 Elsevier Ltd. All rights reserved.
Introduction * Corresponding author. Tel.: +32 2 627 44 28; fax: +32 2 627 41 41. E-mail addresses: [email protected] (E. Vandervennet), [email protected] (A. Huysseune). 1 Tel.: +32 9 264 52 29; fax: +32 9 264 53 44.
Teeth are common features of most classes of vertebrates. Like most other organs, teeth result from reciprocal epithelio-mesenchymal interactions. In all known species, the development of a tooth follows a general pathway in which several stages
0003–9969/$ — see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.archoralbio.2004.12.001
636
can be recognised (initiation, morphogenesis, cytodifferentiation, matrix deposition, mineralisation and maturation, attachment). As teeth are sometimes the only fossil remains of a species, dental features (shape, number and arrangement) are important characters used in systematics to unravel phylogenetic relationships between (groups of) vertebrates.1,2 The Eretmodini are a tribe of closely related cichlid fishes (Teleostei, Cichlidae) originating from Lake Tanganyika in East Africa.3 The four nominal species of this tribe, Eretmodus cyanostictus Boulenger, 1898, Tanganicodus irsacae Poll, 1950, Spathodus erythrodon Boulenger, 1900 and Spathodus marlieri Poll, 1950, resemble each other morphologically to a large extent. Eretmodines are originally divided into three genera on the basis of a distinctive oral tooth shape: spatulate with a crown and neck region in Eretmodus, cylindrical in Spathodus and conical in Tanganicodus.3—5 Based on mitochondrial DNA analyses, Ru ¨ber et al.6 report the existence of six different genetic lineages within this tribe. Within these lineages the three major types of tooth shape appear to have evolved several times, independently of one another. Individuals from a nominal species are distributed over two or more genetic lineages, implying that they probably belong to separate species. For instance, E. cyanostictus is distributed over lineage C (E. cyanostictus) and A (E. cf. cyanostictus). The former represents the originally described E. cyanostictus species, whereas the latter represents a new species. In order to understand, from a developmental perspective, how such dramatically different tooth shapes are acquired within closely related taxa, it is necessary to study the way in which development of an individual tooth is achieved. In particular we wish to know how the enamel organ folds during the process of morphogenesis of the tooth germ and how deposition of the matrix proceeds. Indeed, in teleosts, the epithelio-mesenchymal interface (i.e. the inner surface of the enamel organ) foreshadows the future tooth shape7, due to the specific way in which the hypermineralised cap (the enameloid) is deposited in teleosts. In the literature on teleost tooth development, few papers cover the entire development of a tooth germ (i.e., from initiation to attachment).8 Most studies deal only with specific aspects of tooth development such as enameloid deposition9—12 or mineralisation.13 Huysseune and Sire14 describe in detail the development of (oral) first-generation teeth in the cichlid Hemichromis bimaculatus. However, it is known that the development of replacement teeth in adult teleosts differs in a number of
E. Vandervennet, A. Huysseune aspects from first-generation teeth.14,15 Therefore, and given that a unique tooth pattern is present in the Eretmodini,16 it was necessary to gain a detailed knowledge of the development of replacement teeth in eretmodines as a basis to study the mechanism that generates different (oral) tooth shapes in closely related species as well as within the life cycle of one species. Here we describe the development of replacement teeth in adult Eretmodus cf. cyanostictus using light microscopical observations on semithin as well as on ground sections, microradiographs and stereo-microscopic observations of both alizarin red S stained and unstained jaws. In addition, we used atomic absorption spectroscopy (AAS) to analyse the pigment observed in erupted teeth of Eretmodus cf. cyanostictus, and in particular to measure their iron content. It has been reported that iron is the main contributor to this pigment.17,18
Material and methods Nine adult specimens of E. cf. cyanostictus (standard length (SL) ranging from 37.0 to 52.5 mm) were collected from Lake Tanganyika (1996). Immediately after capture in the field, the animals were fixed and preserved in 80% ethanol. The dissected oral jaws (mandibles and premaxillaries) of the specimens were refixed in a mixture of 1.5% paraformaldehyde and 1.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) to which 0.1 M EDTA was added for decalcification. The jaws were next postfixed in OsO4, dehydrated in a graded ethanol series, and embedded in epon. Serial cross sections of 2 mm thickness were prepared using glass knives, stained with toluidine blue and observed with a light microscope. To prepare ground sections, the jaws (mandibles and premaxillaries) of an adult E. cf. cyanostictus were dehydrated in a graded ethanol series and embedded in methyl metacrylate. Ground sections of 150 mm thickness were prepared using a Buehler Isomet low speed saw. Microradiographic images of the ground sections were made on a sensitive film by X-ray (20 kV, 8 mA) exposure. To study the jaws in toto, they were dissected out of the head. Some of these were cleared in 1—2% KOH, stained with 0.1% alizarin red S in 0.5% KOH and stored in glycerine, others were studied without clearing or staining. Several tooth germs were dissected out of the jaws and used for observation and photography. To determine the total iron content of the enameloid and dentine part of eretmodine teeth samples, we used atomic absorption spectroscopy. Prior to
Replacement tooth development in adult Eretmodus cf. cyanostictus
analysis, the samples were acidified with concentrated HNO3 to 1%. Analysis was carried out on a Varian 800 Atomic Absorption Spectrophotometer with Zeeman background correction. Because of the technical difficulties encountered, we could analyse only five teeth (one whole functional tooth, the enameloid part of one functional tooth, the dentine part of one functional tooth, one unpigmented tooth germ and one orange-coloured membrane covering the enameloid of an unpigmented tooth germ).
Results Since, in fish, tooth replacement takes continuously place during lifetime, jaws of adult E. cf. cyanostictus showed a considerable number of tooth germs, in all stages of development, from initiation stage, up to attached, fully functional teeth. The distribution and localization of the tooth germs in the jaw was in accordance to the pattern described by Huysseune et al.16 Below, a description will be given of the development of replacement teeth in adult E. cf. cyanostictus.
Stage of initiation The first histomorphological indication of tooth formation in E. cf. cyanostictus that we were able to identify was a local downgrowth of the epithelium (Fig. 1). This downgrowth always budded off from the epithelium at the bottom of the mucosal crypt that surrounds the erupted part of a functional tooth. The epithelial strand in E. cf. cyanostictus did not develop straight downwards. Shortly behind the point of budding the strand bent so that it ran parallel to the oral epithelium for some distance. Then the downgrowth penetrated the jawbone and entered the mesenchyme of the medullary cavity to reach the area where the future tooth germ was about to form. This epithelial strand persisted throughout the development of the tooth germ and connected the developing tooth germ to the mucosal surface.
637
(Fig. 2). At this stage, the ODE was composed of cuboidal to slightly cylindrical cells, whereas the cells of the IDE (the future ameloblasts) were already cylindrical. In several tooth germs, we observed several layers of elongated, spindleshaped cells squeezed between IDE and ODE along the prospective convex side of the enamel organ (Fig. 2). At the time the enamel organ formed, the mesenchyme about to become the dental papilla was already distinguishable from the mesenchyme that surrounded the tooth germ. Whereas the latter consisted of elongated spindle-shaped fibroblastlike cells, the mesenchymal cells of the prospective dental papilla were already clearly differentiated: the cells were more rounded, with a larger nucleus, and more cytoplasm. The cells facing the basal lamina (the future odontoblasts) had already taken a slightly polarized appearance (Fig. 2).
Stage of enameloid matrix formation The first matrix to be deposited was the unmineralised precursor of the enameloid. This matrix was a homogeneous, chromophilic substance, in which fibres or fibre bundles were hardly detectable, and which lacked any cellular prolongations emerging from the odontoblasts (Fig. 3). Deposition of enameloid matrix proceeded in an asymmetrical way. The distal side had a delay in comparison with the mesial side. At this stage, the enamel organ was highly differentiated. The ameloblasts facing this unmineralised matrix were extremely tall cells, of slightly varying height, thus resulting in a wavy appearance of the outline of the ameloblast layer. The nuclei of the ameloblasts were situated in the proximal end of the cell (i.e., turned away from the tooth surface). In contrast, the ODE cells were cuboidal and formed a thin layer that followed the wavy outline of the IDE. Many capillaries adjoined the ODE. At this stage, they started to become entrapped within the folds of the enamel organ. The spindle-shaped cells between IDE and ODE were no longer observed.
Stage of morphogenesis
Stage of enameloid matrix mineralisation and maturation
Where the tooth germ was about to develop, the deep end of the epithelial strand next folded into what is called the enamel organ. From the earliest stages onwards, this enamel organ consisted of two layers; the outer dental epithelium (ODE) and the inner dental epithelium (IDE). Initially, the enamel organ extended deeper into the medullary cavity on the future convex and mesial side of the tooth
The enameloid matrix was deposited centripetally, i.e., from the basal lamina inwards. It formed a blade with a convex and a concave surface. The mineralisation of the enameloid matrix started before its deposition was completed. Fully mineralised enameloid hardly showed any evidence of remains of the organic matrix and did not take up any stain.
638
We observed several stages in the development of the enameloid matrix, possibly reflecting the degree of mineralisation. In the youngest tooth germs with ongoing maturation, the fully mineralised matrix was present at the convex (labial) side only (Fig. 4). Maturation then progressed on both sides of the enameloid blade, but left a thin zone of less mineralised matrix in the centre (Fig. 5). The enamel organ was extensively folded during the mineralisation and maturation stages. Whereas
E. Vandervennet, A. Huysseune
the IDE remained in contact with the tooth surface throughout all stages of odontogenesis, folds of the ODE surrounded blood capillaries. The folding of the enamel organ did not occur to the same extent along both sides of the tooth. Indeed, folding was always more pronounced on the side with more advanced matrix maturation, whereas the side facing the less mineralised matrix showed less pronounced folding. When the enamel organ first became moderately folded, dense chromophilic deposits appeared in the
Figures 1—8 E. cf. cyanostictus: toluidine blue stained semithin sections of tooth germs in different stages of development. (Fig. 1) Initiation stage. Note the epithelial downgrowth (black arrowhead) originating from the reduced enamel organ (black asterisk) of a functional tooth (white asterisk). (Fig. 2) Morphogenesis stage. The enamel organ is composed of an inner dental epithelium (white arrowhead) and an outer dental epithelium (black arrowhead). The enamel organ penetrates deeper into the mesenchyme on the future convex side of the tooth. Along this side the two layers seem to be separated by an epithelial layer (black asterisk). The mesenchymal cells of the prospective dental papilla are indicated by a white asterisk. (Fig. 3) Tooth germ in stage of enameloid matrix deposition (white asterisk); the ameloblasts (black asterisk) are extremely tall polarised cells. Note the presence of capillaries along the outer boundary of the enamel organ (arrows). (Fig. 4) Stage of ongoing enameloid mineralisation. The fully mineralised enameloid (black arrowhead) hardly takes up any stain; only traces of organic matrix are left. The less mineralised enameloid is indicated by the white asterisk. (Fig. 5) Tooth germ in a further stage of enameloid mineralisation. Fully mineralised enameloid (black arrowheads) surrounds a central zone of less mineralised enameloid (black asterisk). A white asterisk indicates the dentine matrix. (Fig. 6) Detail of ameloblasts, showing dense chromophilic deposits (black arrowhead) accumulating against the immature mineralised enameloid matrix (black asterisk). (Fig. 7) Cross-sectioned tooth germ showing details of the dentine matrix. Note a pseudo-epithelial layer of odontoblasts (black arrowhead) lining the pulp cavity (black asterisk). Fibre bundles in the dentine are indicated by white arrowhead. (Fig. 8) Longitudinal section through a newly attached tooth. Black arrowheads indicate the unmineralised borderline between dentine base and attachment bone. Small black arrowheads point to osteoblasts lining the surface of the attachment bone. The tip of the cervical loop is highlighted by black asterisks.
Replacement tooth development in adult Eretmodus cf. cyanostictus
639
Fig. 1. (Continued).
ameloblasts opposite the slightly mineralised matrix. These deposits accumulated in the distal part of the ameloblasts, facing the enameloid (Fig. 6). Next, when the enameloid matrix matured, and the enamel organ was elaborately folded, such deposits were found more dispersed in the ameloblasts (Fig. 5); they continued to be present during the whole maturation stage of the enameloid. In some tooth germs, such deposits were also found either at the boundary line between fully mineralised and less mineralised enameloid matrix, or arranged as two lines within the zone of mineralised enameloid (Fig. 5).
Stage of dentine deposition and mineralisation Once deposition of the unmineralised enameloid matrix was finished, the deposition of the dentine matrix started. It is noteworthy that, as long as enameloid was being deposited, the dental papilla steadily receded in front of the forming enameloid, so that the enameloid cap was deposited between two opposite layers of IDE. In contrast, dentine was deposited at the contact area between the dental papilla and the enameloid. Contrary to the enameloid matrix, to which it superficially resembled, the dentine was characterized by the presence of dis-
tinct fibre bundles (Fig. 7). Although cellular prolongations emanating from the odontoblasts were not obvious in the early phase of dentine deposition, a thicker layer of dentine often showed very distinct and long odontoblastic processes. During dentine deposition, the odontoblasts formed a palisade of high columnar cells arranged in a pseudo-epithelial manner against the matrix (Fig. 7). The rest of the dental papilla consisted of a loose network of mesenchymal cells. Around the forming dentine shaft the enamel organ was no longer folded, capillaries were no longer closely associated with the ODE and the cells of the IDE were again low columnar cells. Deposition of the dentine shaft had usually progressed further along one side of the tooth, indicating that its deposition, like the enameloid deposition, proceeded in an asymmetrical way. In contrast to the enameloid, mineralised dentine retained its ability to stain.
Stage of attachment Beneath the base of the dentine shaft, which ended at the level of the cervical loop tip, an annular bone of attachment was deposited (Fig. 8). The tooth base and the attachment bone were ligamentously connected but did not fuse. The inner surface of the
640
E. Vandervennet, A. Huysseune
attachment bone was lined by osteoblasts, which lied in the prolongation of the odontoblasts of the dentine base. Numerous capillaries invaded the pulp cavity.
Table 1 Iron content expressed as absolute values for a complete functional tooth, its constituting parts (enameloid and dentine, respectively), an unpigmented tooth germ and an isolated membrane (enamel organ?), for E. cf. cyanostictus.
Ground sections
Sample
Fe (mg/tooth)
Functional tooth Enameloid Dentine Non-pigmented tooth Pigmented membrane
0.927 0.597 0.094 0.030 0.087
In ground sections observed in transmitted light, all tooth germs appeared to be fully mineralised, with no evidence for a distinction between mineralised and unmineralised enameloid matrix. This was confirmed by comparing the ground sections with their respective microradiographs. Parts of the dentine shaft near the tooth base remained unmineralised.
Iron deposition in enameloid of E. cf. cyanostictus In live and preserved specimens of E. cf. cyanostictus, the enameloid of the functional teeth had a deep orange colour. On semithin sections, an orange-coloured pigment could be observed in the enameloid of (erupted) functional teeth, but never in the tooth germs. In dissected tooth germs of alizarin red S-stained jaws, the enameloid cap showed a colour gradient which corresponded to the developmental age of the tooth germ: from dark pink (intense alizarin red S uptake) in tooth germs with incipient dentine formation, to dark orange (no alizarin red S uptake) in fully developed teeth (Fig. 9). Remarkably, in tooth germs with starting dentine formation, a thin membrane could be peeled off from the enameloid cap of the tooth germ, which appeared to be loaded with alizarin red S positive granules (Fig. 10). In unstained jaws this membrane was loaded with an orange pigment. Such a pigmented membrane was only observed around tooth germs with an unpigmented (white) or slightly pigmented (light orange) enameloid cap,
but not around tooth germs that were more heavily pigmented (dark orange). The results of the iron content analysis of different (parts of) teeth are shown in Table 1. The enameloid contained more iron than the dentine. A functional (pigmented) tooth had a higher iron content than a (almost) fully developed tooth lacking this orange pigment. The iron content of the pigmented membrane, which could be peeled off the enameloid cap of an E. cf. cyanostictus tooth germ, seemed to be low. It should be noted however, that the membrane for which the iron content was measured, only represents a fraction of the weight of the (parts of the) tooth used.
Discussion The way in which the histological development of an E. cf. cyanostictus tooth proceeds complies roughly with what is known for tooth development in other cichlids, or teleost tooth development in general.19 Some remarkable features observed in eretmodine tooth development nevertheless deserve particular attention and will be discussed below. They include the localization and shape of the epithelial down-
Figures 9 and 10 (Fig. 9) Series of dissected tooth germs, prepared from alizarin red S stained jaws of E. cf. cyanostictus, arranged according to an increasing degree of development and mineralisation, from incipient dentin formation (left) to fully formed tooth (right). (Fig. 10) Alizarin red S stained membrane peeled off the enameloid cap of a tooth crown in adult E. cf. cyanostictus similar in appearance to the left tooth germ shown in Fig. 9.
Replacement tooth development in adult Eretmodus cf. cyanostictus
growth, the transient presence of a layer intermediate between IDE and ODE, the close association of capillaries with the enamel organ, the presence of pigment, the asymmetric shape of the enamel organ, the pattern of mineralisation, and finally, the steady withdrawal of the pulp cavity. The tooth germs always budded off from the epithelium at the bottom of the mucosal crypt surrounding the erupted predecessor. We never observed an epithelial strand deriving directly from the oral epithelium. It was recently proposed that this localization, which is also observed in other teleosts, might be related to the presence of a stem cell niche possibly required for the formation of replacement teeth.20 Remarkably, the epithelial strand in E. cf. cyanostictus followed a wavy course prior to penetrating the medullary cavity of the dentary bone. This may be related to the arrangement of the teeth that results in the particular replacement pattern that characterizes eretmodines. Indeed, in eretmodines, the dentition is composed of subsequent oblique rows of teeth, called tooth groups, and is characterized by betweengroup replacement.16 This pattern implies the simultaneous functionality of different teeth belonging to the same tooth family. Hence, predecessor and replacement tooth overlap in their time window of functionality, and therefore the replacement tooth cannot attach at the same position as its predecessor. The replacement tooth needs to develop at some distance labial from its predecessor, a situation possibly enabled by the displacement of the downgrowth of the epithelial strand connecting predecessor and replacement tooth. In most Actinopterygii studied, the enamel organ consists of two layers, the IDE and ODE, which are immediately superimposed onto each other.19 There are nevertheless reports of intermediate cell layers between the IDE and the ODE.10,21—23 The composition of the intermediate layers shows variations among different groups of fish. Sasagawa10 reports that a number of epithelial cells, resembling those of the stellate reticulum in higher vertebrates, are temporarily present between IDE and ODE cells in the cichlid Tilapia nilotica (Oreochromis niloticus) at the middle stage of formation of enameloid matrix. We show here that in E. cf. cyanostictus too, such group of cells was temporarily present during enameloid matrix formation. It seems likely that these cells are involved in the mineralisation process. In E. cf. cyanostictus, as described for other Actinopterygii,13,24,25 capillaries became entrapped in the folds of the enamel organ during the mineralisation and maturation stages, thereby creating a highly folded outer surface of the enamel organ.
641
During these stages, the dental epithelial cells (both IDE and ODE) are involved in the removal of the organic matrix from the enameloid and in the process of mineralisation.26 The ODE cells themselves are involved in the active transport of inorganic ions (calcium, fluoride, iron, . . .) from the capillaries to the IDE.26 The iron is concentrated by the IDE cells as ferritin. This ferritin is digested in ferritin-containing vesicles to release the iron pigment to the enamel.17 In Tilapia the concentration of iron in the IDE cells starts to increase with the appearance of pigment in the same epithelial cells.17 In dissected but unstained jaws of adult E. cf. cyanostictus, we observed a membrane loaded with an orange pigment covering the enameloid cap of tooth germs in the phase of dentine matrix formation (thus, during enameloid mineralisation and maturation stages). The pigmented membrane was only present on teeth that still had an unpigmented or slightly pigmented cap. Later, the enameloid cap became deep orange and a pigmented membrane was no longer detectable. We suspect that this membrane represents the (inner) dental epithelium and that the pigment is a precursor of the iron-containing pigment to be deposited in the enameloid. Iron analysis of eretmodine teeth (Table 1) supports this view. Depending on when the iron deposition into the developing enameloid starts, fish teeth either have iron distributed throughout the entire enameloid layer (in the case of an early start), or they possess an iron-rich surface layer (in the case of a late start).27 In Cichlidae, iron is deposited throughout the entire enameloid layer with the highest concentration at the surface or middle layers.27,28 In functional teeth of adult E. cf. cyanostictus, an orange pigment was dispersed throughout the entire enameloid layer. In light microscopic sections, dense chromophilic deposits were located first in the distal part of the ameloblasts facing slightly mineralised enameloid. In later stages these deposits were dispersed in the ameloblasts (Figs. 5 and 6). These deposits were present from the beginning of the mineralisation stage. Later they also appeared in the enameloid matrix. The order of appearance and localisation of these chromophilic deposits fits our observations for the orange iron pigment. Therefore we suggest that the chromophilic depositions, visualized with toluidine blue dye, correspond to the pigment seen around and in the teeth of unstained jaws, as described above. By the time the enameloid matrix deposition had started, the enamel organ extended deeper into the jaw medullary cavity on the prospective convex surface and on the prospective mesial side of the enameloid cap. Later, the invasion of capillaries, and thus
642
the formation of folds within the enamel organ, started on the convex surface of the tooth as well. Also a delay in matrix deposition on one side of the tooth was observed. Together, these observations suggest that tooth development proceeds asymmetrically. These findings are highly relevant for our interpretation of the acquisition of tooth shape differences between species of eretmodines. Microradiographs of ground sections did not reveal differences in the degree of mineralisation of the enameloid cap. The staining difference observed in the enameloid matrix in toluidine blue stained sections is therefore likely not due to a difference between unmineralised and mineralised matrix, but probably reflects differences in the maturation of the enameloid. Mineralisation of the enameloid starts before its matrix deposition is completed and proceeds in a way which contrasts with other species for which descriptions are available (Cichlasoma,13 Tilapia,29 Hoplognathus30,31,32). In these species, enameloid mineralisation starts when the future outline of the enameloid matrix is completed. In the species listed above, mineralisation always starts at the enameloid-dentine boundary. In Hoplognathus fasciatus, further completion of mineralisation takes place from the surface of enameloid inwards, while in Tilapia nilotica and in Cichlasoma cyanoguttatum, both cichlids, the mineralisation spreads out centrifugally in the cap enameloid. The remarkably different pattern of mineralisation in E. cf. cyanostictus is possibly related to the steady withdrawal of the dental papilla, resulting in the deposition of the enameloid matrix between the two opposing IDE layers, and leaving no pulp cavity in the fully formed enameloid cap. In conclusion, our observations have revealed that the development of an adult E. cf. cyanostictus tooth corresponds to a large extent to what is known for teleost oral tooth development in general (place of initiation of replacement tooth, downgrowth of an epithelial strand, development of the enamel organ within the medullary cavity of the bone). Some features appeared to be related particularly to the eretmodine pattern of the dentition (direction of downgrowth of the epithelial strand), and still others to the particular shape of the teeth (features of the enamel organ, pattern of enameloid mineralisation and maturation and withdrawal of the dental papilla). This description has laid the basis for a detailed study of the enamel organs in the closely related species of eretmodines, aiming to unravel how, from a developmental viewpoint, distinctly different tooth shapes can be generated in such closely related species. The observation of an asymmetrical enamel organ is significant in this respect. Compar-
E. Vandervennet, A. Huysseune
ison with the conical shape of adult T. cf. irsacae and juvenile E. cf. cyanostictus, should allow us to present a model that shows how the folding of the enamel organ changes between conical and spatulate teeth.
Acknowledgments The authors thank Dr. J.Y. Sire and Mrs. M.H. Deschamp (CNRS, Universite ´ Paris 6) for the microradiographs and Prof. Dr. R. Blust and Drs. I. Komjarova (Laboratory for Ecophysiology, Biochemistry and Toxicology, University of Antwerp) for the iron analysis. The research performed by Els Vandervennet is supported by a grant from the Belgian Federal Science Policy Office. Our research further benefited by a grant from the ‘Fonds voor wetenschappelijk onderzoek — Vlaanderen’ (3G010999) and the ‘Bijzonder onderzoeksfonds’ from Ghent University (011B2396).
References 1. Greenwood PH. The haplochromine fishes of the east african lakes. Mu ¨nchen: Kraus International Publications; 1981 (p. 1—839). 2. Thenius E. Za ¨hne und Gebiss der Sa ¨ugetiere. In: Niethammer J, Schliemann H, Starck D, editors. Handbuch der zoologie. Berlin: Walter de Gruyter; 1989. 3. Poll M. Classification des cichlidae du Lac Tanganika: Tribus, ´m Sci 1986;45:1—163. genres et espe `ces. Acad Roy Belg-Me 4. Liem KF. Modulatory multiplicity in the feeding mechanism in cichlid fishes, as exemplified by the invertebrate pickers of Lake Tanganyika. J Zool Lond 1979;189:93—125. 5. Wautier K, Huysseune A, Verheyen E. Tooth shape differences analyzed by biometric and morphometric approaches; a case study on two morphologically very similar lacustrine cichlid species. Connect Tissue Res 2002;43:103—8. 6. Ru ¨ber L, Verheyen E, Meyer A. Replicated evolution of trophic specialisations in an endemic cichlid fish lineage from Lake Tanganyika. Proc Natl Acad Sci USA 1999;96:10230—5. 7. Shellis RP. The role of the IDE in the formation of the teeth in fish. In: Butler PM, Joysey KA, editors. Development, function and evolution of teeth. London: London Academic Press; p. 31—42. 8. Huysseune A, Sire JY. Structure and development of the teeth in three armoured catfish, Corydoras aeneus, C. arcuatus and Hoplosternum littorale (Siluriformes, Callichthyidae). Acta Zool 1997;78:69—84. 9. Shellis RP, Miles AE... Observations with the electron microscope on enameloid formation in the common eel (Anguilla anguilla; Teleostei). Proc R Soc Lond (Biol) 1976;194:253— 69. 10. Sasagawa I. Fine structure of tooth germs during the formation of enameloid matrix in Tilapia nilotica, a teleost fish. Archs Oral Biol 1995;40:801—14. 11. Prostak K, Skobe Z. Ultrastructure of the dental epithelium and odontoblasts during enameloid matrix deposition in cichlid teeth. J Morphol 1986;187:159—72.
Replacement tooth development in adult Eretmodus cf. cyanostictus
12. Prostak K, Seifert P, Skobe Z. Enameloid formation in two tetraodontiform fish species with high and low fluoride contents in enameloid. Arch Oral Biol 1993;38:1031—44. 13. Prostak K, Skobe Z. Ultrastructure of the dental epithelium during enameloid mineralisation in a teleost fish. Cichlasoma cyanoguttatum. Arch Oral Biol 1986;31:73—85. 14. Huysseune A, Sire JY. Structure and development of firstgeneration teeth in the cichlid Hemichromis bimaculatus (Teleostei, Cichlidae). Tissue and Cell 1997;29:679—97. 15. Sire JY, Davit-Beal T, Delgado S, Van Der Heyden C, Huysseune A. First-generation teeth in nonmammalian lineages: evidence for a conserved ancestral character. Microsc Res Tech 2002;59:408—34. 16. Huysseune A, Ru ¨ber L, Verheyen E. A single tooth replacement pattern generates diversity in the dentition in cichlids of the tribe Eretmodini, endemic to Lake Tanganyika (Teleostei: Cichlidae). Belg J Zool 1999;129:157—74. 17. Sasagawa I. Iron accumulation in the dental epithelial cells of Tilapia nilotica, a teleost. ACBTE 1993;3:19—28. 18. Garant P. Observations on the ultrastructure of the ectodermal component during odontogenesis in Helostoma temmincki. Anat Rec 1970;166:167—88. 19. Huysseune A, Sire JY. Evolution of patterns and processes in teeth and tooth-related tissues in nonmammalian vertebrates. Eur J Oral Sci 1998;106:437—781. 20. Huysseune A, Thesleff I. Continuous tooth replacement: the possible involvement of epithelial stem cells. BioEssays 2004;26:665—71. 21. Kerr T. Development and structure of some actinopterygian and urodele teeth. Proc R Soc Lond B 1960;194:253— 69. 22. Peyer B. Comparative odontology. Chicago: University of Chicago Press; 1968 p. 1—347. 23. Ishiyama M, Inage T, Simokawa H. A immunocytochemical study of amelogenin proteins in the developing tooth enamel of the garpike, Lepisosteus oculatus (Holostei, Actinopterygii). Arch Histol Cytol 1999;62:191—7.
643
24. Sasagawa I, Ishiyama M. Fine structure and Ca-ATPase activity of the stratum intermedium cells during ondontogenesis in gars, Lepisosteus Actinopterygii. Connect Tissue Res 2002; 43:505—8. 25. Sasagawa I, Ishiyama M. The features of enameloid formation during odontogenesis in teleosts. In: Mayhall JT, Heikkinen T, editors. Dental Morphology 1998 Proceedings of the 11th international symposium on dental morphology Oulu University Press; 1999. p. 285—92. 26. Sasagawa I. Fine structure of cap enameloid and of the dental epithelial cells during enameloid mineralisation and early maturation stages in the tilapia, a teleost. J Anat 1997;190: 589—600. 27. Suga S, Taki Y. Fluoride and iron concentrations in the teeth of primitive bony fishes (Osteichthyes). In: Suga S, Watabe N, editors. Hard tissue mineralization and demineralization. Tokyo, Hong Kong: Springer Verlag; 1992. p. 17—32. 28. Suga S, Taki Y, Wada K, Ogawa M. Evolution of fluoride and iron concentrations in the enameloid of fish teeth. In: Suga S, Nakahara H, editors. Mechanisms and phylogeny of mineralization in biological systems. Tokyo: Springer Verlag; 1991. p. 439—46. 29. Sasagawa I. Features of the cap enameloid in the pharyngeal teeth of tilapia, a teleost. ACBTE 1995;4:7—14. 30. Isokawa S, Tsubouchi M, Aoki K, Imai M, Kawai A, Tsuchida S. Studies on the developing enameloid of a fish (Hoplognathus fasciatus) I. Mineralization pattern of enameloid matrix. J Nihon Univ Sch Dent 1970;12:43—9. 31. Isokawa S, Yoshida M, Ikeda K, Yamaguchi K, Chikazawa N, Fujioka S. Tooth enameloid formation of a fish (Hoplognathus fasciatus) chronological relationship of matrix formation and mineralization of the enameloid and dentin. J Nihon Univ Sch Dent 1972;14:1—11. 32. Inoue Y, Uchino S, Asano T, Hasegawa S, Fujioka Y. The mineralisation pattern of developing fish tooth enameloid using X-ray historadiography and radioactive tracer (45Ca). J Nihon Univ Sch Dent 1972;15:52—7.